| 00:07 | <Yuhong> | http://logbot.glob.com.au/?c=freenode%23whatwg&s=8+Aug+2012&e=8+Aug+2012#c712060 |
| 00:24 | <Hixie> | http://lists.w3.org/Archives/Public/www-talk/1992MayJun/0022.html is amusing (i'm sure people have run into it before) |
| 00:26 | <Hixie> | http://lists.w3.org/Archives/Public/www-talk/1992JulAug/0017.html too |
| 00:26 | <TabAtkins> | Hixie: Your use-case for <font> was "the only element that allowed 'style'". When people pushed back and you made 'style' global, you killed <font> too. |
| 00:26 | <Hixie> | ah yeah |
| 00:27 | <Hixie> | that sounds right |
| 00:27 | <Hixie> | oh right, <font> was the element that only generators could use or something |
| 00:27 | <TabAtkins> | Yeah. |
| 03:53 | <Hixie> | anyone got IE handy? |
| 03:53 | <Hixie> | any version will do |
| 04:03 | <kennyluck> | Hixie, I have IE9 here. |
| 04:07 | <Hixie> | what do you get on http://damowmow.com/playground/demos/document-write-and-scripts/002.html ? |
| 04:07 | <Hixie> | (and what colour is the text) |
| 04:09 | <kennyluck> | Hixie, text: "0 undefined 1" color: black |
| 04:09 | <Hixie> | wtf |
| 04:09 | <Hixie> | thanks |
| 04:10 | <Hixie> | same as firefox 3.6 |
| 04:10 | <Hixie> | but different than webkit/opera (0 1 2) and different than modern firefox (0 2 2) |
| 04:10 | <Hixie> | i wonder if it's not blocking load on the style sheet or something |
| 04:12 | <Hixie> | kennyluck: if you're still there, can you try http://damowmow.com/playground/demos/document-write-and-scripts/002-long.html ? |
| 04:12 | <Hixie> | it'll dump a bunch of stuff |
| 04:12 | <Hixie> | if i could get you to /msg it to me that'd be awesome |
| 04:14 | <Hixie> | ohhhh |
| 04:14 | <Hixie> | i know what's wrong |
| 04:14 | <Hixie> | hang on |
| 04:15 | <mthz> | is anyone aware of a copy of the html5lib tokenizer test data that is in alignment with the current spec? |
| 04:15 | <mthz> | the current tests are a mess, especially w/r/t error handling for entities |
| 04:15 | <Hixie> | kennyluck: can you try http://damowmow.com/playground/demos/document-write-and-scripts/002.html in IE now? |
| 04:15 | <Hixie> | kennyluck: (i'd forgotten to take out the console.log() calls which IE and old Firefoxen don't support) |
| 04:15 | <kennyluck> | Hixie, sure. 002-long.html gave http://pastebin.mozilla.org/1779848 |
| 04:16 | <mthz> | or, does anyone which tests the following page even refer to (i.e. where can i get them?) http://wiki.whatwg.org/wiki/Parser_tests |
| 04:16 | <Hixie> | mthz: that refers to the tests in the code.google.com repo for html5lib |
| 04:16 | <kennyluck> | Hixie, text "2 1 1" color: black |
| 04:17 | <Hixie> | kennyluck: !! |
| 04:17 | <zewt> | it's pretty (something) that even today you have to jump hoops to see console.log output on ios |
| 04:17 | <Hixie> | kennyluck: (you sure it's black and not dark blue?) |
| 04:17 | <Hixie> | kennyluck: "2 1 1" is typical IE behaviour. complete nonsense. :-P |
| 04:17 | <kennyluck> | Hixie, muh, I can't quite tell. Can you use another color or something. |
| 04:17 | <Hixie> | don't worry about it |
| 04:17 | <Hixie> | it's not critical |
| 04:18 | <Hixie> | oh... i see what it's doing |
| 04:18 | <Hixie> | interesting |
| 04:18 | <kennyluck> | Hixie, oh yeah. It's navy. |
| 04:18 | <Hixie> | k, cool, thanks |
| 04:21 | <mthz> | Hixie: Are you aware of a copy of those tests that mirrors the spec? They look way out of date. It even looks like the validator.nu code mimcs this behavior |
| 04:21 | <Hixie> | do you have an example of something out of date? |
| 04:21 | <Hixie> | i'm sure they're not perfectly up to date, but my understanding was that those were the latest tests |
| 04:22 | <mthz> | Yeah, or maybe I'm just misreading the spec.. I'll paste |
| 04:22 | <mthz> | Test: Entity in attribute without semicolon ending in i (test1.test): Input=<h a='¬i'> |
| 04:23 | <mthz> | It expects that to yield an error presumably b/c the matched entity didn't end with ';' and has ascii after the last match |
| 04:23 | <mthz> | but the spec doesn't say to emit an error |
| 04:23 | <mthz> | bunch of entity related tests like that that all expect similar errors when my reading of the spec doesn't suggest there should be any |
| 04:24 | Hixie | looks at the spec |
| 04:25 | <Hixie> | (in general i wouldn't be surprised if the error count was more out of date that the expected DOM tree, fwiw) |
| 04:25 | <mthz> | i'm only running the tokenizer tests -- not the tree builder -- but yes, i agree. I think all of the out-datedness I've seen is related to error counts |
| 04:26 | <Hixie> | what does it say the output should be for that test? |
| 04:27 | <mthz> | "output":["ParseError", ["StartTag", "h", {"a":"¬i"}]]}, |
| 04:27 | <mthz> | I think that's correct save for the parse error |
| 04:27 | <Hixie> | yeah, i think you're right |
| 04:27 | <Hixie> | do you know when the test was updated? |
| 04:27 | <Hixie> | that part of the spec was changed at some point |
| 04:27 | <Hixie> | to make it not an error |
| 04:28 | <mthz> | sep 24 2009 ;-) |
| 04:28 | <mthz> | makes sense then |
| 04:29 | <mthz> | I guess I'll compile a list of the ones I come across and log a bug for html5lib |
| 04:29 | <Hixie> | while you're at it, add a test for "¬=" (also not a parse error and not treated as an entity) if it's not there already |
| 04:30 | <Hixie> | that was changed april 2010 |
| 04:30 | <Hixie> | looks like as of 2009 it was already not a parse error though |
| 04:31 | <mthz> | unfortunately I can't update the html5lib code itself -- probably a bad idea to update the tests and not fix the code |
| 04:31 | <Hixie> | well if the test is wrong the test is wrong :-) |
| 04:31 | <mthz> | heh |
| 04:32 | <mthz> | fwiw, not sure if you saw above, but Henry Sivonen's validator.nu code shows the same vehavior for all of the false negatives i've found |
| 04:32 | <mthz> | i'll send them mail |
| 04:32 | <mthz> | thanks hixie |
| 04:32 | <MikeSmith> | mthz: you know that's actually the same code that Firefox uses |
| 04:32 | <MikeSmith> | html parser code |
| 04:33 | <mthz> | Mike: the code is semantically correct -- it's merely error reporting which i don't think ever surfaces in the browser |
| 04:33 | <MikeSmith> | ah |
| 04:33 | <MikeSmith> | yeah, true |
| 04:33 | <Hixie> | ah i was wrong, the change was in r4960, also april 2010 |
| 04:33 | <Hixie> | the change to make it not a parse error i mean |
| 04:34 | <mthz> | MikeSmith: random question, but do/did you work for google? i think we may have met before =) |
| 04:34 | <Hixie> | and r4959 was the revision that made = act like a-z in attributes |
| 04:34 | <MikeSmith> | mthz: there's a different Mike Smith who was a product manager for Chrome |
| 04:34 | <Hixie> | mthz: anyway, in any e-mail you said please let them know r4959 and r4960, to make it easier for them to track it |
| 04:35 | <mthz> | mike: ahh -- that's the one |
| 04:35 | <mthz> | hixie: will do.. thanks |
| 04:35 | <Hixie> | thank _you_! |
| 04:35 | <mthz> | for my own reference -- what repository do those revisions refer to? |
| 04:35 | <Hixie> | whatwg |
| 04:35 | <Hixie> | svn.whatwg.org/web-apps |
| 04:35 | <mthz> | figured.. thanks |
| 04:35 | <mthz> | later all |
| 04:35 | <Hixie> | later |
| 04:39 | <Hixie> | why do Opera and Chrome (not Safari) return 0 for document.createElement('canvas').getContext('2d').moveTo.length |
| 04:39 | <Hixie> | Firefox and Safari get it right |
| 04:39 | <Hixie> | (2) |
| 04:47 | <Hixie> | ok the change i just made to the spec is for a ridiculously complicated edge case. it's crazy that that is web-compat-critical. |
| 04:47 | <heycam> | O_o |
| 04:48 | <Hixie> | it's specifically for the case of an inline <script> that executes while the parser is re-entrantly parsing due to a document.write() call, in the case of there being a <link rel=styleesheet> pending that is itself blocking script execution. |
| 04:49 | <Hixie> | and it controls whether or not that nested inline script blocks or not. |
| 04:49 | <Hixie> | s/controls/affects/ |
| 04:49 | <heycam> | oh, sorry I thought you specced the moveTo.length being 0 |
| 04:49 | <Hixie> | the HTML parser, i mean |
| 04:49 | <Hixie> | heycam: no, no, that'd be crazy. |
| 04:49 | <heycam> | yes :) |
| 04:49 | <zewt> | yeah. *that* would be crazy |
| 04:49 | <zewt> | *cough* |
| 04:49 | <Hixie> | :-P |
| 05:48 | <annevk> | Does anyone know what hg-git is called in MacPorts? |
| 05:56 | <MikeSmith> | annevk: py27-hggit |
| 05:56 | <annevk> | Thanks, I ended up using easy_install |
| 05:59 | <zcorpan> | Hixie: opera gets .length wrong (is 0) all over the place. known bug, low prio. |
| 06:00 | <Hixie> | it makes feature detection for new features that consisted of adding an optional argument harder |
| 06:00 | <Hixie> | if that helps bump up the prio |
| 06:03 | <annevk> | MikeSmith: how do you authenticate for github? |
| 06:03 | <MikeSmith> | ssh |
| 06:03 | <annevk> | MikeSmith: I have the hg push ssh+git:// thingie, but how do I pass username/password? |
| 06:05 | <annevk> | ah found it |
| 06:06 | <jgraham> | Hixie: Does anyone apart from you try to do feature detection in that way? I imagine not if Chrome and Opera both get it wrong |
| 06:07 | <Hixie> | if opera and chrome didn't get it wrong, it would be a good way to do it for e.g. the new arcTo() |
| 06:07 | <Hixie> | but yeah, i don't think anyone actually does it |
| 06:09 | <jgraham> | annevk: (I don't recommend the hg push ssh+git method, but if it worked for MikeSmith maybe it's fine) |
| 06:10 | <MikeSmith> | jgraham, annevk : what I'm using for the current mirroring to github/w3c is this: |
| 06:10 | <MikeSmith> | hg bookmark -d master |
| 06:10 | <MikeSmith> | hg bookmark -fr default master |
| 06:10 | <MikeSmith> | hg gexport |
| 06:10 | <MikeSmith> | git push github master |
| 06:11 | <jgraham> | Yes, I was going to say that hg gexport worked better for me |
| 06:11 | <jgraham> | I just had to look up what it was |
| 06:11 | <annevk> | my main problem is authentication |
| 06:11 | <annevk> | the rest seems to work fine |
| 06:11 | <jgraham> | Hmm, but for github authentication should be key-baed |
| 06:11 | <jgraham> | *key-based |
| 06:12 | <jgraham> | You shouldn't need to add username or pw anywhere |
| 06:12 | <zcorpan> | annevk: http://krijnhoetmer.nl/irc-logs/whatwg/20120827#l-771 |
| 06:13 | <jgraham> | That is, if you set up the right keys on your computer per the instructions on github |
| 06:13 | <jgraham> | Then do what MikeSmith said up to the last but one step |
| 06:13 | <annevk> | github says https is recommended |
| 06:13 | <jgraham> | Then git remote add origin git⊙gc:whatwg/dom.git |
| 06:13 | <jgraham> | git push -u origin master |
| 06:14 | <jgraham> | Since I just copied/pasted that bit from the github page I'm sure it's fine |
| 06:20 | <divya> | jgraham: !!1 just wanted to check if you are on track to be in paris |
| 06:20 | <divya> | for the event |
| 06:20 | <divya> | and hoping you could give a talk too while at it |
| 06:21 | <jgraham> | divya: Yes, and yes |
| 06:22 | <divya> | jgraham: sweet i will put you in touch with someone who wanted to find out what you would be talking aobout. hurray. |
| 06:25 | <jgraham> | divya: Cool |
| 06:25 | Ms2ger | approves |
| 06:25 | <divya> | :)) Ms2ger do you ever make an IRL appearance? |
| 06:25 | <divya> | MikeSmith: would you also be at TPAC? |
| 06:26 | <MikeSmith> | hey divya |
| 06:26 | <MikeSmith> | yeah, will be there |
| 06:26 | <divya> | so also for Test the Web Forward then MikeSmith ? |
| 06:26 | <divya> | :PPP |
| 06:26 | <divya> | you should ideall |
| 06:26 | <divya> | y |
| 06:26 | <Ms2ger> | Ever? Maybe |
| 06:27 | <divya> | ahaha :) |
| 06:28 | <MikeSmith> | divya: what's the dates? |
| 06:28 | <divya> | MikeSmith: 26-27 oct |
| 06:28 | <divya> | just before TPAC |
| 06:28 | <divya> | hold on it should be on csswg wiki |
| 06:29 | <divya> | MikeSmith: http://wiki.csswg.org/test/events/paris-2012 |
| 06:29 | <divya> | MikeSmith: this is the internal organizational thing we will have the website ready by end of the week |
| 06:29 | <MikeSmith> | I see |
| 06:30 | <jamesr_> | othermaciej: i believe the wiki proposal for unrelated context does address your question about noreferrer |
| 06:31 | <jamesr_> | othermaciej: http://wiki.whatwg.org/wiki/Links_to_Unrelated_Browsing_Contexts#Current_Usage_and_Workarounds |
| 06:31 | <Ms2ger> | divya, s/Glassman/Glazman/? |
| 06:31 | <divya> | ahaha Ms2ger yes >_> |
| 06:31 | <othermaciej> | I see |
| 06:32 | <othermaciej> | jamesr_: is there a difference other than whether the Referer header is sent? |
| 06:32 | <divya> | Ms2ger: the person who has been updating the wiki doesnt know the members >_> i will have to dig out my credentials from the dark corners of my email archive |
| 06:32 | <Ms2ger> | Oh dear :) |
| 06:32 | <jamesr_> | assuming noreferrer nulls out the window.opener, etc, i don't think so |
| 06:33 | <othermaciej> | jamesr_: also is it really ok to drop support for the window.opener behavior of noreferrer? that seems bold |
| 06:34 | <jamesr_> | you mean with rel=noreferrer and target=_blank ? |
| 06:34 | <divya> | Ms2ger: or glazou can edit it himself :PP |
| 06:34 | <divya> | Ms2ger: i can assure you though he is pretty sharp and planned the first one. |
| 06:34 | <jamesr_> | we've been doing it in chromium for several years, i don't know if i was around to know if we got compat issues when we started doing that |
| 06:35 | <othermaciej> | jamesr_: yes, the requirement to null out window.opener as per http://www.whatwg.org/specs/web-apps/current-work/multipage/links.html#link-type-noreferrer |
| 06:35 | <othermaciej> | jamesr_: the wiki section you cited seems to suggest that this behavior of rel=noreferrer could be removed |
| 06:35 | <jamesr_> | oh i see. i don't know about that |
| 06:35 | <othermaciej> | which seems all kinds of wrong to me |
| 06:35 | <jamesr_> | i thought you were referring to the nulling out behavior, which i think is fine |
| 06:36 | <othermaciej> | yes, the nulling behavior is good and I think a logical corollary of not sending referrer |
| 06:36 | <othermaciej> | and also it's probably relied on by more than just gmail |
| 06:36 | <jamesr_> | oh i'm sure |
| 06:36 | <jamesr_> | i can't find anything in the proposal suggesting that noreferrer would stop doing this |
| 06:38 | <othermaciej> | oh, I misread |
| 06:38 | <annevk> | zcorpan: so yeah it was the caching :/ |
| 06:38 | <annevk> | guess I'll just empty the cache again |
| 06:38 | <othermaciej> | I was looking at "Google Chrome also has a non-standard trick for opening links in a new process by using window.open(), setting the resulting window's opener to null, and then navigating the new window to a different site" |
| 06:39 | <jamesr_> | ah gotcha |
| 06:39 | <jamesr_> | yeah that's a way to get the unrelated behavior with window.open() |
| 06:39 | <jamesr_> | as opposed to a link |
| 06:39 | <zcorpan> | annevk: is it possible to prune the cache automatically (or empty it once a every few months automatically)? |
| 06:39 | <jamesr_> | that's also a behavior other pages probably use now |
| 06:39 | <jamesr_> | (and one that not-so-infrequently breaks with gmail, heh) |
| 06:40 | <othermaciej> | does that method result in sending Referer or no? |
| 06:40 | <othermaciej> | I can't tell |
| 06:40 | <annevk> | zcorpan: I guess, but I'm fine with this |
| 06:40 | <othermaciej> | the page says gmail uses that technique, but I would be surprised if a mail client wanted to send referer |
| 06:40 | <othermaciej> | as that seems like a privacy issue |
| 06:40 | <annevk> | zcorpan: I'd rather we remove the cache, but maybe that would be too much of a hit on the SVN server |
| 06:40 | <jamesr_> | depends on what it is, you can sanitize the URL with a redirect |
| 06:41 | <jamesr_> | i'll bet if you asked charlie reis he'd know the answer to all of these sorts of questions. this particular behavior is tricky to implement and i'm not super familiar |
| 06:41 | <othermaciej> | I'm in fact trying to ask him questions via email on the whatwg list |
| 06:42 | <annevk> | zcorpan: might actually have been a hiccup from the SVN server too |
| 06:43 | <jamesr_> | the proposal does mention supporting rel="…" in general for window.open() would be useful. then you could just set noreferrer and there'd be no question about what happens |
| 06:44 | <othermaciej> | that part of the proposal seems ok to me (though only a small subset of link relations seem like they would be relevant in that usage) |
| 06:45 | <othermaciej> | I'm just not really clear on when you would want the "unrelated" behavior instead of "noreferrer" |
| 06:47 | <othermaciej> | I am puzzled that he implies mail clients would want it, as sending referer with the url of the mail client seems clearly bad |
| 06:53 | <jamesr_> | othermaciej: the use case was google reader |
| 06:53 | <jamesr_> | or imagine some sort of content aggregator app |
| 06:53 | <jamesr_> | seems pretty reasonable to send a referer from my rss reader to the page |
| 06:54 | <othermaciej> | most content aggregators I use don't presume to open windows for me, but if they did, I guess I could see how they'd want to null opener but still send referer |
| 07:46 | <Ms2ger> | annevk, <annevk> please someone remind me tomorrow |
| 07:48 | <jgraham> | We are reminding annevk that it is tomorrow? |
| 07:49 | <zcorpan> | Ms2ger: it's already fixed |
| 07:49 | <Ms2ger> | Ah, good |
| 07:49 | <Ms2ger> | Also, nobody ever uses function.length |
| 07:53 | <jgraham> | Hixie: http://damowmow.com/playground/demos/document-write-and-scripts/002.html in Opera gives me 0 2 2 |
| 08:40 | <jgraham> | So, if you document.write in the load event, how many history entries should you end up with? Spec/firefox/IE say 2 (one before the write, one after), Opera/WebKit say 1 (after the write) |
| 08:41 | <jgraham> | Is there any reason the spec behaviour is more useful? AFAICT it usually breaks the back button |
| 08:42 | <jgraham> | <script>onload = function() {document.write("Go back and this will be rerun, so you will end up in the same place")}</script> |
| 08:48 | <zcorpan> | Ms2ger: that's why it's low prio :-) |
| 09:05 | <gsnedders> | Ms2ger: Test suites use Function.length! |
| 09:07 | <zcorpan> | gsnedders: that's why it's a known bug :-P |
| 10:58 | <zcorpan> | so, w3c -= annevk |
| 11:03 | <beverloo> | his new company just needs to join as a member :p |
| 11:04 | <beverloo> | it's disturbing nonetheless |
| 11:05 | <jgraham> | It's not surprising |
| 11:07 | <jgraham> | Perhaps W3C will eventually learn that their craziness is bad if it has consequences |
| 11:07 | <jgraham> | Although it didn't work so well with HTML |
| 11:08 | <jgraham> | With the whole fork -> unfork -> fork thing |
| 11:53 | <withub> | Is there any way to get the noreferrer relation to work on a script tag? |
| 11:54 | <withub> | so that when the src of the script tag is fetched from the remote server, a referer won't be set? |
| 11:54 | <withub> | s/set/sent |
| 11:56 | <annevk> | beverloo: the Member Agreement is not much better |
| 14:13 | <jgraham> | Hixie: Is there a difference between the behaviour in bug 18459 and just updating the document base to be the new document base on pushState unless there is a base element in which case not updating it? |
| 14:14 | <jgraham> | s/to be the new document base/to be the new document address/ |
| 16:03 | <dglazkov> | good morning, Whatwg! |
| 17:21 | <Hixie> | jgraham: changing the <base> element dynamically presumably also affects things |
| 17:22 | <Hixie> | jgraham: (re document.write, weird, i guess that changed at some point? i must have an older opera at home) |
| 17:34 | <annevk> | Hixie: just to be clear, you prefer html.spec.whatwg.org over what we have today? |
| 17:34 | <annevk> | Hixie: I guess that works too |
| 17:35 | <annevk> | Hixie: it's a little bit more hassle with the DreamHost people and becomes semi-problematic if editors change |
| 17:35 | <Hixie> | it's mostly because that gives the editors complete control over the subdomain, and i don't have to worry about managing anything or worrying about security |
| 17:35 | <Hixie> | why is there any hassle with dreamhost? |
| 17:36 | Hixie | has 60+ subdomains, and is pretty used to it :-P |
| 17:36 | <annevk> | there was last time I think, because you assigned the subdomain directly to me |
| 17:36 | <Hixie> | changing editors is a concern, true, but the solution then is _don't change editors_ :-P pick up a responsibility and keep it for life :-P |
| 17:36 | <Hixie> | oh right, yeah, if we do that it's a bit more complex |
| 17:36 | <Hixie> | i figured we'd just have users under my account |
| 17:37 | <TabAtkins> | "The spec editor mates for life with their chosen spec." |
| 17:38 | <Ms2ger> | TabAtkins, fortunately we don't require picking a single spec... |
| 17:38 | <Hixie> | annevk: anyway, if it's all in a github repo it's not a huge deal if the editors change, you just wipe the directory and have hte new one do their stuff afresh :-) |
| 17:38 | <TabAtkins> | Nor do we require a single editor per spec. It's a happy polyamarous family. |
| 17:38 | <annevk> | Hixie: fair enough |
| 17:39 | Hixie | is of the opinion that if you can be editing multiple specs, your spec isn't big enough, and multiple editors means diluted blame, which is even worse :-P |
| 17:39 | <TabAtkins> | The relationship graph of editors and specs is K2, luckily. |
| 17:39 | <TabAtkins> | Hixie: I edit one spec, it's just split up into multiple documetns. |
| 17:39 | <Hixie> | hah |
| 17:48 | <jgraham> | Hixie: Yeah, changing the base is an interesting case I guess |
| 17:52 | <Hixie> | any opera people around with opinions on <template>? |
| 17:53 | <Hixie> | as in http://dvcs.w3.org/hg/webcomponents/raw-file/tip/spec/templates/index.html#introduction |
| 17:53 | <Hixie> | trying to work out whether to put it in the html spec or not |
| 17:53 | <TabAtkins> | Hixie: The only person we know of who's cared about it from Opera is jgraham. |
| 17:53 | <jgraham> | Hixie: Something like http://hoppipolla.co.uk/tests/pushState/003.html ? |
| 17:54 | <annevk> | well "cared" would also include me :p |
| 17:54 | <TabAtkins> | At least, in mailing lists so far. |
| 17:54 | <TabAtkins> | But you're not Opera, so there. |
| 17:54 | <jgraham> | Hixie: What sort of opinion are you looking for here? |
| 17:54 | <Hixie> | (i hate the way pushState() tests make it hard to view source in some browsers :-P) |
| 17:54 | <annevk> | TabAtkins: :) |
| 17:54 | <zewt> | i'm doing templating-in-markup already; i just stick my templates in a @hidden block, then clone its contents into a DocumentFragment |
| 17:54 | <Hixie> | jgraham: go/no-go, or anything more detailed |
| 17:55 | <TabAtkins> | zewt: That's problematic if you want to avoid loading resources. :/ |
| 17:55 | <zewt> | doesn't have nice magical syntax for filling in contents, but that's easy enough with querySelector |
| 17:55 | <jgraham> | I don't think that we object to the idea of web components or templates, but we might of course want to argue about the details |
| 17:55 | <zewt> | TabAtkins: i guess; in my case I do want everything preloaded |
| 17:56 | <jgraham> | So no objection to adding it to the HTML spec if that is generally considered to be the best home for it |
| 17:56 | <TabAtkins> | zewt: Okay. More generally, if you're stamping out templated html multiple times, you may want to vary a resource location in each stamp, like <img src='image{i}'> or whatever. |
| 17:56 | <zewt> | TabAtkins: my first instinct is that you should be able to (somehow) say something like "@hidden and also don't load resources for this hidden element", though of course that's far more complicated than @hidden today |
| 17:56 | <jgraham> | (if that was a vauge answer, it was a vauge question :p) |
| 17:56 | <Hixie> | jgraham: ok |
| 17:57 | <zewt> | i do that, eg. var newInstance = copyTemplate(template); newInstance.querySelector("img.some-image").src = image[i] |
| 17:57 | <Hixie> | jgraham: any objection in particular to the proposed parsing mechanism (parsing "children" of <template> into an anonymous separate document)? |
| 17:57 | <zewt> | like I said it's not as pretty as a more thorough templating system might be able to make it, but it's been working pretty well |
| 17:59 | <TabAtkins> | zewt: That's basically what <template> does. ^_^ |
| 17:59 | <Hixie> | jgraham: yeah, that test is exactly what i meant |
| 17:59 | <Hixie> | jgraham: very interesting |
| 17:59 | <TabAtkins> | (re: hidden and don't load resources) |
| 17:59 | <TabAtkins> | (also, don't allow querySelector and friends to match into it) |
| 17:59 | <zewt> | TabAtkins: seems like it might be a useful thing in general, even without templates |
| 18:00 | <jgraham> | Hixie: I don't have a strong objection to the magical parsing. I am aware that hsivonen does and I think his argument has merit, but I don't have an alternative design for the feature that is better |
| 18:00 | <Hixie> | jgraham: chrome and firefox are freakishly interoperable on that test given the lack of basis for their behaviour in any spec |
| 18:01 | <Hixie> | jgraham: k |
| 18:01 | <dglazkov> | Hixie: you may want to check with Tony Russ and Rafael Weinstein who have been actively working on this spec first. |
| 18:01 | <jgraham> | Hixie: I hate things that are freakishly interopable but contradict all known specs :) |
| 18:02 | <Hixie> | dglazkov: i am in close contact with rafael |
| 18:02 | <dglazkov> | Hixie: great! |
| 18:02 | <dglazkov> | how close? :P |
| 18:02 | <Hixie> | he's one of the only people i've had a video conf call with this year :-P |
| 18:02 | <dglazkov> | whoa |
| 18:02 | <Hixie> | possibly in fact the only person :-P |
| 18:02 | <Hixie> | more than once! |
| 18:03 | <dglazkov> | btw, we spoke with other Mozillians (bz, sicking), and they don't think the magic parsing design is bad. |
| 18:03 | <dglazkov> | but raf would have a much more detailed information on that meeting |
| 18:04 | <Hixie> | k |
| 18:04 | <Hixie> | (hopefully they can let me know that directly, too) |
| 18:04 | <dglazkov> | I think raf is planning to send another mail to the list about that |
| 18:04 | <Hixie> | k |
| 18:26 | <JVoracek> | |
| 19:39 | <annevk> | since I put my email at the top of the MetaExtensions page instead of getting emails about registering new users people complain about validation errors and wonder how to fix those |
| 19:39 | <annevk> | I might leave remove my email address again Hixie and only deal with those people that ask for an account on IRC |
| 19:40 | <annevk> | heh, scumbag steve |
| 19:41 | <TabAtkins> | annevk: Yeah, change the message to be "come to IRC and ask" |
| 19:42 | <zewt> | put a 4-page algorithm for deriving your email address |
| 19:45 | <annevk> | TabAtkins: done |
| 19:51 | <charlvn> | annevk: so you still unemployed? |
| 19:51 | <annevk> | charlvn: gonna start doing some work next week, but no pay for now |
| 19:52 | <annevk> | no need for money at the moment |
| 19:52 | <charlvn> | taking a break every now and then is a good thing |
| 19:52 | <charlvn> | enjoy it while it lasts :) |
| 19:52 | <annevk> | it's been a good two months so far :) |
| 19:53 | <charlvn> | nice |
| 19:54 | <annevk> | charlvn: are you actually in the Netherlands these days? |
| 19:54 | <charlvn> | annevk: haven't been outside of the EU in almost a year |
| 19:54 | <annevk> | should visit one of those Fronteers meetups one day |
| 19:55 | <charlvn> | i have been putting it off but it's a good idea |
| 19:55 | <charlvn> | thanks for reminding me |
| 19:55 | <annevk> | and time it appropriately so we can meet, would love to learn some more Afrikaans :) |
| 19:55 | <charlvn> | lol my afrikaans has gone down the drain :S |
| 19:56 | <annevk> | too bad, together with Belgian I think it's more Dutch than Dutch |
| 19:56 | <charlvn> | well they like to translate things "properly", not just take english words over |
| 19:57 | <charlvn> | actually i find it rather painful |
| 19:57 | <annevk> | right, Dutch is borrowed all over |
| 19:57 | <charlvn> | when i try to use software in afrikaans the number of wtfs per minute is pretty high |
| 19:57 | <charlvn> | i don't understand half of the application anymore |
| 19:57 | <annevk> | and even made less Dutch (kado -> cadeau) |
| 19:57 | <annevk> | haha |
| 19:58 | <annevk> | admittedly I only use English software |
| 19:58 | <charlvn> | at work we have this application, the dutch translation pack has been bought from belgium |
| 19:58 | <charlvn> | none of us use it, we immediately switch to english |
| 19:59 | <charlvn> | we just talk of a "policy", not a "beleid" so it confuses us :) |
| 20:04 | <divya> | hey whatwg what is jgraham 's email id |
| 20:04 | divya | is lazy to check mailing list. |
| 20:04 | <gsnedders> | divya: For Opera? jgraham . |
| 20:04 | <divya> | gsnedders: yes. o thnx. |
| 20:04 | <divya> | jgraham@o i suppose |
| 20:04 | <gsnedders> | Yeah. |
| 20:05 | <divya> | thnx gsnedders |
| 20:06 | <gsnedders> | divya: PS: I am now your LDAP directory. |
| 20:07 | <gsnedders> | *not |
| 20:07 | <annevk> | not hers, but for #whatwg you are :p |
| 20:08 | <gsnedders> | :( |
| 20:10 | <jgraham> | gsnedders: You have many things in common with LDAP |
| 20:10 | <jgraham> | Like no one actually understands how you work |
| 20:13 | <jgraham> | divya: I got your email, thanks |
| 20:13 | <divya> | gsnedders: :))))) |
| 20:13 | <divya> | jgraham: AHAHAHA poor gsnedders |
| 20:14 | <jgraham> | I guess you are expecting me to reply |
| 20:14 | <divya> | but it is true you are mysteriously magical gsnedders MUCH LIKE LDAP |
| 20:14 | <divya> | jgraham: not immediately no. email is not Instant Messaging :P |
| 20:14 | <divya> | or Text or call. |
| 20:14 | divya | wonders how many media of communication exist |
| 20:15 | <jgraham> | divya: You are not expecting me to reply at all, or I should wait some unspecified time to make it sufficiently async to fit your notion of what is appropriate for email? :) |
| 20:15 | <divya> | jgraham: REPLY WHEN YOU PLEASE |
| 20:15 | <divya> | jgraham: or whenever is convenient!!! |
| 20:16 | divya | takes note of jgraham's penchant for nitpicking |
| 20:16 | <annevk> | not a fan of dubstep, but http://www.youtube.com/watch?v=WgII2gDY-Rw is hilarious |
| 20:17 | <jgraham> | dglazkov: Is this your "implement Web Components or die" expression? https://lh4.googleusercontent.com/-utqdXXy_BGU/UDuu7l702AI/AAAAAAAAKKM/-ptCs0_sffU/s657/IMG_20120827_101442.jpg |
| 20:17 | <annevk> | especially the woman in pink is great |
| 20:17 | <dglazkov> | jgraham: the only response to this that I can have is: https://lh4.googleusercontent.com/-utqdXXy_BGU/UDuu7l702AI/AAAAAAAAKKM/-ptCs0_sffU/s657/IMG_20120827_101442.jpg |
| 20:18 | <dglazkov> | or it's emoticon equivalent, <_< |
| 20:18 | <divya> | ahahahaha dglazkov |
| 20:19 | <divya> | annevk: is this the video you have been recommending since morning |
| 20:19 | <annevk> | your morning, my early evening :) |
| 20:19 | <annevk> | just passing on the greatness showed to me by robbert |
| 20:21 | <divya> | that lady in pink is bestest |
| 20:25 | <gsnedders> | jgraham: I work in the same manner as most people in this industry: by hitting keys on a keyboard. |
| 21:01 | <Hixie> | annevk: weird, i've only ever had people asking for accounts |
